This invention relates to polypeptides that are resistant to thermal degradation. More particularly, this invention relates to thermally stabilized insulin analogues. Even more particularly, this invention relates to insulin analogues that are chemically and thermally stabilized by the incorporation of the element fluorine, chlorine, or bromine into an amino acid of the insulin analogue. These elements are classified as halogens and are distinguished from the ordinary constituents of proteins by their atomic radius, electronegativity, stereoelectronic distribution of partial charges, transmitted effects on the stereoelectronic properties of neighboring atoms in a molecule.
The engineering of ultra-stable proteins, including therapeutic agents and vaccines, may have broad societal benefits in regions of the developing world where electricity and refrigeration are not consistently available. An example of a therapeutic protein susceptible to thermal degradation is provided by insulin. The challenge posed by its chemical and physical degradation is deepened by the pending epidemic of diabetes mellitus in Africa and Asia. Because chemical degradation rates of insulin analogs correlate inversely with their relative stabilities, the design of ultra-stable formulations may enhance the safety and efficacy of insulin replacement therapy in such regions.
The utility of some halogen substitutions in amino acids is well established in medicinal chemistry. Indeed, fluorinated functional groups are critical to the efficacy of such widely prescribed small molecules as atorvastatin (Liptor™), an inhibitor of cholesterol biosynthesis, and fluoxetine hydrochloride (Prozac™), a selective serotonin reuptake inhibitor used in the treatment of depression and other affective disorders. Although the atomic radius of fluorine is similar to that of hydrogen, its large inductive effects modify the stereo-electronic properties of these drugs, in turn enhancing their biological activities. Such observations have motivated the study of fluorinated amino acids in proteins. Similar considerations of physical organic chemistry pertain to the incorporation of larger halogen atoms, such as chlorine and bromine. The small molecule montelukast sodium (Singulair™) is a leukotriene inhibitor whose pharmaceutical properties are enhanced by covalent incorporation of a chlorine atom.
Attention has previously focused on the use of multiply fluorinated aliphatic side chains (such as trifluoro-γ-CF3-Val, trifluoro-δ-CF3-Val, trifluoro-δ-CF3-Ile, hexafluoro-γ1,2-CF3-Val, and hexafluoro-δ1,2-CF3-Leu) to maximize the gain in hydrophobicity associated with this modification. An example is provided by the stabilization of a model α-helical motif, the homodimeric coiled coil. Its interfacial aliphatic chains were simultaneously substituted by trifluoro-analogs, creating a fluorous core whose stability is enhanced by 0.3-2.0 kcal/mole. The degree of stabilization per fluorine atom is <0.1 kcal/mole. More marked stabilization per fluorine atom has been achieved in an unrelated α-helical domain by substitution of a single internal Phe by pentafluoro-Phe (F5-Phe)1 (ΔΔGu 0.6 kcal/mole per five fluorine atoms). Stabilization occurs only at one specific position in the protein, suggesting that its mechanism requires a particular spatial environment. The structure of the F5-Phe-modified domain is identical to that of the unmodified domain. Structural stabilization however, is only a portion of the requirements for a biologically active polypeptide. At least a significant portion of the activity must also be maintained.
An extensive literature describes the use of fluorine labels in proteins as 19F-NMR probe. Whereas such labels are widely regarded as non-perturbing, applications of fluorinated amino acids in protein engineering seek to exploit their altered physicochemical properties. Studies of a model α-helical fold (the villin headpiece subdomain) containing single F5-Phe substitutions have demonstrated that whether and to what extent such a modification may affect protein stability depends in detail on structural environment. Indeed, the stability of this model fold was enhanced by an F5-Phe substitution only at one of the seven sites tested in the core. It should be noted however, that this stabilizing effect was only demonstrated in a nonstandard polypeptide analogue containing a sulfur atom in the main chain near the site of fluorination; in our hands the self-same F5-Phe substitution in villin headpiece subdomain with native polypeptide main chain has no effect on its stability. Thus, to our knowledge, bona fide data demonstrating enhancement of protein stability due to halogenation of an aromatic residue in an otherwise native protein has not previously been reported. These observations suggest that a generally hydrophobic environment does not in itself assure that the modification will be stabilizing. Because protein interiors are often stabilized by aromatic-aromatic interactions with a specific distance- and angular dependence, a subset of stabilizing F5-Phe substitutions may adopt particularly favorable perfluoroaryl/aryl geometries. Such interactions arise from the asymmetric distribution of partial charges in these aromatic systems. Modulation of the chemical, physical, and biological properties of proteins by the site-specific incorporation of chlorine or bromine atoms into modified amino acids are less well characterized in the scientific literature than are the above effects of incorporation of fluorine atoms.
Aromatic side chains may engage in a variety of weakly polar interactions, involving not only neighboring aromatic rings but also other sources of positive- or negative electrostatic potential. Examples include main-chain carbonyl- and amide groups in peptide bonds.
Administration of insulin has long been established as a treatment for diabetes mellitus. Insulin is a small globular protein that plays a central role in metabolism in vertebrates. Insulin contains two chains, an A chain, containing 21 residues and a B chain containing 30 residues. The hormone is stored in the pancreatic β-cell as a Zn2+-stabilized hexamer, but functions as a Zn2+-free monomer in the bloodstream. Insulin is the product of a single-chain precursor, proinsulin, in which a connecting region (35 residues) links the C-terminal residue of B chain (residue B30) to the N-terminal residue of the A chain (FIG. 1A). Although the structure of proinsulin has not been determined, a variety of evidence indicates that it consists of an insulin-like core and disordered connecting peptide (FIG. 1B). Formation of three specific disulfide bridges (A6-A11, A7-B7, and A20-B19; FIGS. 1A and 1B) is thought to be coupled to oxidative folding of proinsulin in the rough endoplasmic reticulum (ER). Proinsulin assembles to form soluble Zn2+-coordinated hexamers shortly after export from ER to the Golgi apparatus. Endoproteolytic digestion and conversion to insulin occurs in immature secretory granules followed by morphological condensation. Crystalline arrays of zinc insulin hexamers within mature storage granules have been visualized by electron microscopy (EM).
Insulin functions in the bloodstream as a monomer, and yet it is the monomer that is believed to be most susceptible to fibrillation and most forms of chemical degradation. The structure of an insulin monomer, characterized in solution by NMR, is shown in FIG. 1D. The A-chain consists of an N-terminal α-helix (residues A1-A8), non-canonical turn (A9-A12), second α-helix (A12-A18), and C-terminal extension (A19-A21). The B chain contains an N-terminal arm (B1-B6), β-turn (B7-B10), central α-helix (B9-B19), β-turn (B20-B23), β-strand (B24-B28), and flexible C-terminal residues B29-B30. The two chains pack to form a compact globular domain stabilized by three disulfide bridges (cysteines A6-A11, A7-B7, and A20-B19).
Absorption of regular insulin is limited by the kinetic lifetime of the Zn-insulin hexamer, whose disassembly to smaller dimers and monomers is required to enable transit through the endothelial lining of capillaries. The essential idea underlying the design of Humalog® and Novolog® is to accelerate disassembly. This is accomplished by destabilization of the classical dimer-forming surface (the C-terminal anti-parallel β-sheet). Humalog® contains substitutions ProB28→Lys and LysB29→Pro, an inversion that mimics the sequence of IGF-I. Novolog® contains the substitution ProB28→Asp. Although the substitutions impair dimerization, the analogs are competent for assembly of a phenol- or meta-cresol-stabilized zinc hexamer. This assembly protects the analog from fibrillation in the vial, but following subcutaneous injection, the hexamer rapidly dissociates as the phenol (or m-cresol) and zinc ions diffuse away. The instability of these analogs underlies their reduced shelf life on dilution by the patient or health-care provider. It would be useful for an insulin analogue to augment the intrinsic stability of the insulin monomer while retaining the variant dimer-related β-sheet of Humalog®.
Use of zinc insulin hexamers during storage is known and represents a classical strategy to retard physical degradation and chemical degradation of a formulation in the vial or in the reservoir of a pump. Because the zinc insulin hexamer is too large for immediate passage into capillaries, the rate of absorption of insulin after subcutaneous injection is limited by the time required for dissociation of hexamers into smaller dimers and monomer units. Therefore, it would advantageous for an insulin analogue to be both (a) competent to permit hexamer assembly at high protein concentration (as in a vial or pump) and yet (b) sufficiently destabilized at the dimer interface to exhibit accelerated disassembly—hence predicting ultra-rapid absorption from the subcutaneous depot. These structural goals walk a fine line between stability (during storage) and instability (following injection).
Amino-acid substitutions in insulin have been investigated for effects on thermodynamic stability and biological activity. No consistent relationship has been observed between stability and activity. Whereas some substitutions that enhance thermodynamic stability also enhance binding to the insulin receptor, other substitutions that enhance stability impede such binding. The effects of substitution of ThrA8 by several other amino acids has been investigated in wild-type human insulin and in the context of an engineered insulin monomer containing three unrelated substitutions in the B-chain (HisB10→Asp, ProB28→Lys, and LysB29→Pro) have been reported. Examples are also known in the art of substitutions that accelerate or delay the time course of absorption. Such substitutions (such as AspB28 in Novalog® and [LysB28, ProB29] in Humalog®) can be and often are associated with more rapid fibrillation and poorer physical stability. Indeed, a series of ten analogs of human insulin have been tested for susceptibility to fibrillation, including AspB28-insulin and AspB10-insulin. All ten were found to be more susceptible to fibrillation at pH 7.4 and 37° C. than is human insulin. The ten substitutions were located at diverse sites in the insulin molecule and are likely to be associated with a wide variation of changes in classical thermodynamic stability. Although a range of effects has been observed, no correlation exists between activity and thermodynamic stability.
Insulin is a small globular protein that is highly amenable to chemical synthesis and semi-synthesis, which facilitates the incorporation of nonstandard side chains. Insulin contains three phenylalanine residues (positions B1, B24, and B25). Conserved among vertebrate insulins and insulin-like growth factors, the aromatic ring of PheB24 packs against (but not within) the hydrophobic core to stabilize the super-secondary structure of the B-chain. PheB24 lies at the classical receptor-binding surface and has been proposed to direct a change in conformation on receptor binding.
The present theory of protein fibrillation posits that the mechanism of fibrillation proceeds via a partially folded intermediate state, which in turn aggregates to form an amyloidogenic nucleus. In this theory, it is possible that amino-acid substitutions that stabilize the native state may or may not stabilize the partially folded intermediate state and may or may not increase (or decrease) the free-energy barrier between the native state and the intermediate state. Therefore, the current theory indicates that the tendency of a given amino-acid substitution in the insulin molecule to increase or decrease the risk of fibrillation is highly unpredictable.
Fibrillation, which is a serious concern in the manufacture, storage and use of insulin and insulin analogues for diabetes treatment, is enhanced with higher temperature, lower pH, agitation, or the presence of urea, guanidine, ethanol co-solvent, or hydrophobic surfaces. Current US drug regulations demand that insulin be discarded if fibrillation occurs at a level of one percent or more. Because fibrillation is enhanced at higher temperatures, diabetic individuals optimally must keep insulin refrigerated prior to use. Fibrillation of insulin or an insulin analogue can be a particular concern for diabetic patients utilizing an external insulin pump, in which small amounts of insulin or insulin analogue are injected into the patient's body at regular intervals. In such a usage, the insulin or insulin analogue is not kept refrigerated within the pump apparatus and fibrillation of insulin can result in blockage of the catheter used to inject insulin or insulin analogue into the body, potentially resulting in unpredictable blood glucose level fluctuations or even dangerous hyperglycemia. At least one recent report has indicated that lispro insulin (an analogue in which residues B28 and B29 are interchanged relative to their positions in wild-type human insulin; trade name Humalog®) may be particularly susceptible to fibrillation and resulting obstruction of insulin pump catheters.
Insulin fibrillation is an even greater concern in implantable insulin pumps, where the insulin would be contained within the implant for 1-3 months at high concentration and at physiological temperature (i.e. 37° C.), rather than at ambient temperature as with an external pump. Additionally, the agitation caused by normal movement would also tend to accelerate fibrillation of insulin. In spite of the increased potential for insulin fibrillation, implantable insulin pumps are still the subject of research efforts, due to the potential advantages of such systems. These advantages include intraperitoneal delivery of insulin to the portal circulatory system, which mimics normal physiological delivery of insulin more closely than subcutaneous injection, which provides insulin to the patient via the systemic circulatory system. Intraperitoneal delivery provides more rapid and consistent absorption of insulin compared to subcutaneous injection, which can provide variable absorption and degradation from one injection site to another. Administration of insulin via an implantable pump also potentially provides increased patient convenience. Whereas efforts to prevent fibrillation, such as by addition of a surfactant to the reservoir, have provided some improvement, these improvements have heretofore been considered insufficient to allow reliable usage of an implanted insulin pump in diabetic patients outside of strictly monitored clinical trials.
As noted above, the developing world faces a challenge regarding the safe storage, delivery, and use of drugs and vaccines. This challenge complicates the use of temperature-sensitive insulin formulations in regions of Africa and Asia lacking consistent access to electricity and refrigeration, a challenge likely to be deepened by the pending epidemic of diabetes in the developing world. Insulin exhibits an increase in degradation rate of 10-fold or more for each 10° C. increment in temperature above 25° C., and guidelines call for storage at temperatures <30° C. and preferably with refrigeration. At higher temperatures insulin undergoes both chemical degradation (changes in covalent structure such as formation of iso-aspartic acid, rearrangement of disulfide bridges, and formation of covalent polymers) and physical degradation (non-native aggregation and fibrillation.
Amino-acid substitutions have been described in insulin that stabilize the protein but augment its binding to the insulin receptor (IR) and its cross-binding to the homologous receptor for insulin-like growth factors (IGFR) in such a way as to confer a risk of carcinogenesis. An example known in the art is provided by the substitution of HisB10 by aspartic acid. Although AspB10-insulin exhibits favorable pharmaceutical properties with respect to stability and pharmacokinetics, its enhanced receptor-binding properties were associated with tumorigenesis in Sprague-Dawley rats. Although there are many potential substitutions in the A- or B chains that can be introduced into AspB10-insulin or related analogues to reduce its binding to IR and IGFR to levels similar to that of human insulin, such substitutions generally impair the stability of insulin (or insulin analogues) and increase its susceptibility to chemical and physical degradation. It would be desirable to discover a method of modification of insulin and of insulin analogues that enabled “tuning” of receptor-binding affinities while at the same time enhancing stability and resistance to fibrillation. Such applications would require a set of stabilizing modifications that reduce binding to IR and IGFR to varying extent so as to offset the potential carcinogenicity of analogues that are super-active in their receptor-binding properties.
Therefore, there is a need for alternative insulin analogues, including those that are stabilized during storage while maintaining at least a portion of the biological activity of the analogue.